Towards simplification of hydrologic modeling: identification of dominant processes

• Published in: Hydrology and earth system sciences Vol. 2016; no. 1; p. 1
• Main Authors: Markstrom, S. L, Hay, L. E, Clark, M. P
• Format: Journal Article
• Language: English
• Published: Copernicus GmbH 25.01.2016
• Subjects: Analysis; Hydrology; Precipitation (Meteorology); Runoff; Soil moisture
• ISSN: 1027-5606; 1607-7938

An application of the Precipitation-Runoff Modeling System, a distributed-parameter hydrologic model, has been developed for the conterminous United States. In this study, two different aspects of the complexity in applying this model has been addressed: (1) the number of input parameters and (2) the interpretation of model output. Parameter sensitivity analysis was used to simplify the application of the hydrologic model through identification of parameters related to dominant hydrologic processes (baseflow, evapotranspiration, runoff, infiltration, snowmelt, soil moisture, surface runoff, and interflow) at various geog raphic scales. These processes correspond to variables for which objective functions (mean, autoregressive lag 1, and coefficient of variation) are computed. Categories of parameter sensitivity were developed in various ways, on the basis of geographic location, hydrologic process and model response. Visualization of these categories provide insight into model performance and useful information about how to structure the modeling application to take advantage of as much local information as possible. The results of this study indicates that (1) the choice of objective function and output variables have a strong influence on parameter sensitivity, (2) the dimensionality of distributed-parameter hydrology models can be reduced by removing input parameters, output variables and objective functions from consideration on the basis of selection by hydrological process, (3) different hydrological processes require different numbers of parameters for simulation, and (4) some model sensitive parameters influence only one hydrologic process, while others may influence many.